ocr/extract

Extract layout-aware text and tables from a scanned PDF or image.

Dynamic (cost in response)

charged on success

What it does

Runs OCR over a scanned PDF or image and returns the recognized content as markdown per page, preserving structure including tables.

Primary use cases

Reading scanned contracts, receipts, or forms that have no text layer.
Converting photographed documents into editable, searchable text.
Pulling tables out of scanned reports as markdown.

Why use this tool

OCR quality and layout/table reconstruction are hard to do locally. Mistral OCR returns clean, structured markdown rather than a flat character dump.

Good to know

For PDFs that already contain a text layer, pdf-tools/extract_text is cheaper. Use pages to limit which pages are processed (billing is per page processed).

Parameters

pagesarrayoptional

Optional 0-indexed page numbers to process. Defaults to all pages.

outputstringoptionaldefault: "markdown"

Preferred output style. The response always includes per-page markdown.

input_urlstringrequired

Pre-signed GET URL of the scanned PDF or image.